| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Immunology Department, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Mexico City
Correspondence: Dr. Carlos Rosales, Department of Immunology, Instituto de Investigaciones Biomédicas—UNAM, Apto. Postal 70228, Cd. Universitaria, México D.F.—04510, Mexico. E-mail: carosal{at}servidor.unam.mx
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONFcR for IgG (Fc{gamma}R)SIGNAL TRANSDUCTION DURING...ENCLOSURE OF THE PHAGOCYTIC...PHAGOSOME MATURATIONCONCLUSIONREFERENCES |
|---|
Key Words: neutrophil • macrophage • monocyte • immunoglobulin
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONFcR for IgG (Fc{gamma}R)SIGNAL TRANSDUCTION DURING...ENCLOSURE OF THE PHAGOCYTIC...PHAGOSOME MATURATIONCONCLUSIONREFERENCES |
|---|
Phagocytosis is triggered by the interaction of opsonins, which cover the particle to be internalized, with specific receptors on the surface of the phagocyte. These receptors include the Fc receptors (FcR), which bind to the Fc portion of immunoglobulins (Ig) [5 ], and the complement receptors [6 ], which bind to complement on opsonized particles. Progressive interaction of the receptors with their ligands allows phagocytosis to proceed in a "zipper-like" manner until complete internalization of the particle is achieved within a specialized structure, the phagosome. The phagosome then travels inside the cell to fuse with lysosomes [7 ] and in this way becomes a microbicidal organelle [3 ].
Biochemical regulation of phagocytosis in the different phagocytes is now a very active field of research. Most work regarding the regulation of phagocytosis has been done on FcR and complement receptors, although other receptors, such as CD14, the lipopolysaccharide receptor; lectins, which recognize specific sugar residues on the surface of microorganisms; and scavenger receptors, which recognize altered lipids, sugars, or proteins on the surface of apoptotic cells, are also capable of mediating phagocytosis [8 , 9 ].
Over the last couple of years, a tremendous amount of new data has been presented on different aspects of the phagocytic process. Most of the new information relates to the biochemical events that regulate phagocytosis via FcR. From this information, it has become evident that FcR-mediated phagocytosis occurs as a series of steps that are regulated in a nonlinear manner [10 11 12 ]. In this review, we will focus on this new information, and we will present a general model of our present understanding of the signal transduction events that regulate phagocytosis mediated by FcR.
|
FcR for IgG (Fc R)
|
|---|
|
TOPABSTRACTINTRODUCTIONFcR for IgG (Fc{gamma}R)SIGNAL TRANSDUCTION DURING...ENCLOSURE OF THE PHAGOCYTIC...PHAGOSOME MATURATIONCONCLUSIONREFERENCES |
|---|
R and FcR for IgE (Fc
R) and IgA (Fc
R) have been described, and they are all members of the Ig superfamily of receptors to which the T cell receptors (TCR) and B cell receptors belong [13
]. Interaction of FcR with their Ig ligands triggers a wide series of leukocyte responses that play crucial roles in inflammation and immunity. These responses include phagocytosis, antibody-dependent cell-mediated cytotoxicity (ADCC), release of proinflammatory mediators, and production of cytokines [13
, 14
].
Three classes of FcR have been recognized to date: FcRI, FcRII, and FcRIII. Each class of FcR consists of several individual receptor isoforms [5
]. FcRI (CD64) is coded by three different genes (A, B, and C) and is expressed in monocytes, macrophages, and interferon--stimulated neutrophils. FcRI is a high-affinity receptor capable of binding monomeric IgG. FcRII (CD32) is also coded by three different genes (A, B, and C) and is a low-affinity receptor capable of binding only multimeric IgG. FcRIIA is expressed mainly in phagocytic cells but also in natural killer (NK) cells. FcRIIB is constitutively expressed in T and B lymphocytes. In phagocytes, the mRNA for FcRIIB has been known to exist for some years [15
16
17
], but only recently the protein product of this transcript was conclusively demonstrated in human monocytes [18
]. In these cells, FcRIIB expression appears to be up-regulated by interleukin-4 treatment [18
, 19
]. FcRIII (DC16) is coded by two genes (A and B) and is also a low-affinity receptor. FcRIIIA is expressed in macrophages and in monocytes at a lower level, whereas FcRIIIB is expressed exclusively in neutrophils. In contrast to FcRIIIA and all the other FcR, FcRIIIB lacks a transmembrane region and a cytoplasmic tail. This receptor is anchored to the membrane by a glycophosphatidylinositol (GPI) moiety [5
].
Most FcR isoforms, including FcRI [20
], FcRIIA [21
], and FcRIIIA [22
], are able to mediate phagocytosis. In contrast, FcRIIB negatively regulates phagocytosis [18
, 23
]. The neutrophil isoform FcRIIIB is capable of inducing calcium signaling [24
] and actin polymerization [25
], but its role in phagocytosis is still controversial [26
]. Activation of phagocytosis depends on the signals delivered by these receptors after they are cross-linked on the membrane of the phagocyte. Although recent studies on various cell systems demonstrate the complexity of the phagocytic signaling, we now begin to see how the various molecules participate in this process.
|
SIGNAL TRANSDUCTION DURING PHAGOCYTOSIS BY FcR
|
|---|
|
TOPABSTRACTINTRODUCTIONFcR for IgG (Fc{gamma}R)SIGNAL TRANSDUCTION DURING...ENCLOSURE OF THE PHAGOCYTIC...PHAGOSOME MATURATIONCONCLUSIONREFERENCES |
|---|
R is initiated by the clustering of these receptors by IgG-coated particles. This event is followed by phosphorylation of specific tyrosine residues within special amino acid motifs, called ITAMs (immunoreceptor tyrosine-based activation motifs) [27
], located on the cytoplasmic portion of FcRIIA and on accessory or 
chains that associate with FcRI and FcRIIIA. The initial ITAM phosphorylation is caused by enzymes of the Src tyrosine-kinase family [28
]. These kinases remain inactive through the interaction of a phosphorylated tyrosine residue, located near their carboxy end with their own SH (Src homology)-2 domain. This interaction causes these enzymes to fold and block their catalytic sites [29
]. Activation seems to involve dephosphorylation, which in leukocytes, may be carried out by the phosphatase CD45 [30
]. Activation also involves autophosphorylation and may require direct protein-protein interactions [28
]. Apparently a small fraction of active Src kinases is stably associated to the cytoplasmic tails of the ITAM-containing FcR chains so that receptor cross-linking will induce activation of kinases in their vicinity [27
, 31
].
At least six members of the Src family have been identified in phagocytes: Fgr, Fyn, Hck, Lyn, Yes, and Src [28
, 31
, 32
]. Some of them have been found associated to specific receptors [28
]. For example, in monocytes, Hck and Lyn have been found associated with FcRI, whereas only Hck has been found associated with FcRIIA [33
, 34
]. Similarly, in neutrophils, Fgr has been found in association with FcRIIA [35
], and Lck with FcRIIIA in NK cells [36
]. These kinases are important for the early phosphorylation events following FcR engagement [28
]. However, their particular involvement in phagocytosis remains unclear, as Src genetically deficient (Src-/-) macrophages did not present alterations in phagocytosis [37
]. Similarly, phagocytosis by Hck-/- macrophages was normal, although Lyn-/- cells presented a mild defect [38
]. Moreover, Hck-/-, Fgr-/-, and Lyn-/- triple-deficient phagocytes had a more pronounced delay in phagocytosis but still not complete abolition of the process [38
]. Also, it was found that Hck-/- Fgr-/- macrophages still expressed Fyn, Src, and Yes, which could partly account for the residual phagocytic activity. Upon stimulation with IgG-coated particles, these kinases were all activated and redistributed to actin-rich phagocytic cups [32
]. These findings suggest that a high level of redundancy exists for these kinases and that all of them are needed during activation of phagocytosis. However, at least one of them (Fgr) has been reported to have a negative regulatory role on phagocytosis [39
].
Phosphorylated ITAMs then become docking sites for the SH-2 domains of Syk [40
, 41
]. Syk is a tyrosine kinase, expressed in all hematopoietic cells, and related to the kinase Zap70, which plays a fundamental role in TCR signaling [42
]. Docking of Syk triggers its phosphorylation (by Src kinases) and its activation [10
, 43
]. Recruitment and activation of Syk are very important for phagocytosis, as indicated by Syk activation after FcR stimulation [43
], by inhibition of phagocytosis with Syk antisense oligonucleotides [44
], and by failure of macrophages from Syk-/- mice to internalize IgG-opsonized particles [45
]. Similarly, in neutrophils, pharmacological inhibition of Syk abrogates phagocytosis of IgG-coated particles [46
]. However, the exact role of Syk in this process remains unclear. Some reports indicate that Syk is needed for the formation of the actin-filament cup that assembles beneath the opsonized particle during FcR-mediated phagocytosis [32
, 46
47
48
], and others indicate that Syk-/- cells have normal actin assembly but are incapable of particle internalization [45
]. In addition, the fact that several nonhematopoietic cell lines (which do not express Syk) are capable of phagocytosis when expressing the appropriate receptors [49
] indicates that Syk may not be essential for this process. Alternatively, it is also possible that a Syk-like molecule exists in these cells. However, coexpression of Syk and FcR in nonphagocytic cell lines results in more efficient phagocytosis [50
], which underlines the fundamental role of Syk in this process. Variations in the phagocytic potential of various FcR may also be, at least in part, a result of differences in their interaction with Syk at the initial steps of phagocytic signaling [51
]. FcRIIA consistently bound Syk kinase when this receptor retained just one (the most carboxy-terminal) tyrosine within the ITAM, whereas the chain, associated with FcRI and FcRIIIA, required both tyrosines within the ITAM for efficient coimmunoprecipitation with Syk [51
]. In addition, FcRIIA and the chain contain a tyrosine seven amino acids upstream of the ITAM motif. This upstream tyrosine was found to be important for FcRIIA but not for chain signaling to phagocytosis [51
]. Future studies will certainly help us to understand how these differences relate to the various signaling pathways that follow Syk activation.
Downstream signaling pathways
The downstream pathways stimulated by active Syk are incompletely understood. Many signaling molecules, including calcium, protein kinase C (PKC), phospholipase A2 (PLA2), phospholipase C (PLC), phospholipase D (PLD), phosphatidylinositol 3-kinase (PI-3K), extracellular signal-regulated kinase (ERK), and GTPases of the Rho family, have been implicated in phagocytic signaling (Fig. 1
), but their exact roles in this process remain elusive.
|
R engagement, an increase in cytosolic Ca2+ concentration has always been considered relevant for phagocytosis, as all FcR can induce it [24
, 52
, 53
], and as this Ca2+ concentration increase is greatest in the cytoplasm surrounding the phagocytic cup [54
]. Although all FcR can induce an increase in cytosolic Ca2+ concentration, the need for Ca2+ in phagocytosis seems to be associated with the different forms of FcR. Using the murine macrophage P388D1 cell line transfected with several FcR mutants, it was reported that Ca2+-independent phagocytosis is a property associated with the use of -chains by FcRI and FcRIII [55
]. This might indicate differences in signal transduction between the -chain-associated receptors and the ITAM present in FcRIIA.
This simple picture for Ca2+ dependence in phagocytosis gets more complicated when we look at reports of phagocytosis by different professional phagocytes. IgG-mediated phagocytosis by human neutrophils was found to be Ca2+-dependent [56
] and also Ca2+-independent [57
]. In contrast, phagocytosis [58
, 59
] and phagosome-lysosome fusion [60
] in macrophages seem to be Ca2+-independent. Similarly, IgG-mediated phagocytosis by monocytes seems to be independent of Ca2+ [61
]. In addition, a single cell type is also capable of Ca2+-dependent and Ca2+-independent phagocytosis, according to the activation state of the cell [62
]. As all these cell types express more than one type of FcR, these differences may suggest that phagocytes use different FcR for phagocytosis and/or that signaling during phagocytosis may vary in response to signals from other receptors that modify the activation state of the phagocyte. Thus, the exact role of Ca2+ during phagocytosis in different phagocytes remains an open question.
During IgG-mediated phagocytosis, the Ca2+ concentration increase is greatest in the cytoplasm surrounding the phagocytic cup [54
], and it is thought that this Ca2+ is a direct consequence of FcR signaling. However, a recent report indicates that this increase may be caused by the exit of Ca2+ from the phagosome into the cytosol through Ca2+ channels, rather than by Ca2+ released from intracellular stores [63
]. The reduction of Ca2+ concentration in the phagosome seems important for phagosome maturation [63
]. Independently of its origin, Ca2+ seems important for triggering actin depolymerization around phagosomes [64
]. This action may be achieved by activation of gelsolin by a local increase in Ca2+ concentration. Gelsolin caps the barbed end of actin filaments, preventing filament elongation [65
]. Neutrophils from gelsolin-deficient mice showed a serious defect in FcR-mediated phagocytosis [66
]. However, Ca2+-dependent depolymerization of actin filaments around already internalized particles was normal in the same gelsolin-deficient cells [66
], suggesting that Ca2+ is important in other aspects of phagocytosis as well.
Another unresolved issue is the actual second messenger that causes the Ca2+ increase. Inositol trisphosphate (IP3) is the principal second messenger responsible for Ca2+ release from intracellular stores [67
] (Fig. 1)
. However, in neutrophils [24
] and mast cells [68
], the Ca2+ release after FcR activation has been found to be independent of this metabolite. Also, in monocytes, FcRI cross-linking induces an IP3-independent Ca2+ rise [69
]. In addition, indirect evidence suggested that L-plastin, an actin-binding protein that is phosphorylated in response to phagocytosis [70
], might participate in the IP3-independent Ca2+ increase mediated by FcRIIA in neutrophils [71
]. In contrast, in macrophages, it was reported that FcRI or FcRII stimulation induces a PLC-mediated Ca2+ increase, dependent on IP3 production [72
]. Additionally, it was also found that sphingosine-1-phosphate is the actual second messenger responsible for the cytoplasmic Ca2+ increase produced after FcRI cross-linking in mast cells [68
] and FcRI cross-linking in monocytes [69
]. However, when these monocytes are differentiated to a more macrophage phenotype, FcRI leads to PLC activation and a more prolonged Ca2+ response [72
]. Also, it was reported very recently that FcRI triggers not only a fast and transient sphingosine-1-phosphate-mediated Ca2+ release but also a PLC-mediated second, slower wave of Ca2+ release from intracellular stores [73
]. Thus, the relevant second messenger responsible for the cytoplasmic Ca2+ increase used by FcR during phagocytosis may vary according to the particular receptor involved and to the differentiation state of the cell.
PKC
Enzymes of the PKC family comprise a large group of serine/threonine kinases. These enzymes are grouped into four subfamilies based on structure and cofactor requirements: conventional (, ßI, ßII, ), novel (
, , 
, 
), atypical (), and recently described (µ, 
) [74
]. Pharmacological inhibition or expression of dominant negative isoforms of PKC reduced phagocytosis to a great extent in several systems [75
76
77
78
79
]. However, the precise role of the particular PKC isoforms involved in phagocytosis remains unclear. Involvement of PKC activity in complement receptor-mediated phagocytosis has been clearly demonstrated [80
]. In the case of FcR-mediated phagocytosis, data are more complex [80
81
82
]. Differences in these reports regarding the involvement of PKC may be a result of the use of various PKC isoforms in phagocytosis. PKC has been found in macrophage phagosomes during complement receptor 3- and FcR-mediated phagocytosis [77
, 80
, 82
]. Also, PKCß [83
], PKC [84
], PKC [85
], and PKC [82
] have all been reported to accumulate in the phagosome membrane during FcR-mediated phagocytosis. These data suggest that different PKC isoforms are responsible for different aspects of phagocytosis [82
]. The particular PKC isoform involved in phagocytosis may depend on the specific FcR involved. Future studies will determine which PKC isoform is activated by each FcR. Additionally, the PKC isoforms activated by FcR seem to vary depending on the differentiation state of the cell. In U937 monocytes, it was found that FcRI engagement leads to an increase in PKC activity that is Ca2+-independent and corresponds to translocation to the membrane of the PKC isoforms , , and [86
]. In U937-differentiated macrophages, FcRI engagement leads to PKC activity that is Ca2+-dependent and corresponds to membrane translocation of the conventional PKC isoforms , ß, and [84
, 86
].
Despite the reports mentioned above, the precise role of PKC during phagocytosis remains unclear. However, the observation that some of the downstream targets of PKC are important regulators of phagocytosis provides important clues. A PKC-dependent pathway for ERK activation has been described in neutrophils, monocytes, and mouse macrophages [46
, 77
, 79
, 87
] (Fig. 1)
, and ERK has recently been observed to be an important regulator of phagocytosis in neutrophils and macrophages [79
]. Additionally, in monocytes and macrophages, at least one isoform of PLA2 (calcium-independent PLA2) appears to be regulated by PKC [88
, 89
] (Fig. 1)
. Pharmacological inhibition of PLA2 results in phagocytosis arrest [88
]. Thus, PKC may regulate phagocytosis through activation of ERK and some isoforms of PLA2. Plekstrin, the main PKC target in platelets, is also localized to phagosome membranes during FcR-mediated phagocytosis [85
]. MARCKS (myristoylated alanine-rich C kinase substrate) is a known PKC target that cross-links actin filaments [90
]. MARCKS is also found in phagosomes [91
], suggesting it has a role in phagocytosis. However, macrophages from MARCKS-/- mice presented normal phagocytosis [92
]. The identification of the precise PKC isoforms and their targets necessary for regulation of phagocytosis in different phagocytes will certainly clarify our understanding of this important FcR-mediated function.
PLAs
Several PLAs have been reported to be involved in phagocytosis. Although the exact manner in which they participate in this process remains unclear. PLA2 mediates arachidonic acid (AA) release from phosphatidylcholine or phosphatidylethanolamine [93
, 94
]. Leukocytes express three isoforms of PLA2: a secreted, Ca2+-dependent PLA; a cytosolic, Ca2+-dependent PLA (cPLA2); and a cytosolic, Ca2+-independent PLA (iPLA2) [94
]. The participation of PLA2 and AA release in FcR-mediated phagocytosis was demonstrated in monocytes treated with bromoenol lactone, a selective inhibitor of calcium-independent PLA2 [95
, 96
]. Addition of exogenous AA to PLA2-inhibited monocytes restored phagocytosis [88
, 96
, 97
]. AA can be metabolized into proinflammatory mediators through the lipooxygenase and cyclooxygenase metabolic pathways. However, inhibition of cyclooxygenase and lipooxygenase did not affect phagocytosis [95
, 98
]. Thus, these reports suggest that PLA2 participates in phagocytosis through the production of AA, which itself, and not its bioactive metabolites, acts as a second messenger to regulate phagocytosis [94
]. In monocytes, iPLA2 seems to be regulated by PKC [76
, 87
, 97
] (Fig. 2A
). In neutrophils and macrophages, cPLA2 is regulated by ERK and p38 mitogen-activated protein kinase (MAPK) [99
, 100
], and iPLA2 is regulated by PKC [89
] (Fig. 2B)
. Whereas in monocytes, AA release appears to rely only on the PKC-iPLA2 pathway, in neutrophils and macrophages, the PKC-iPLA2 and the ERK/p38-cPLA2 pathways appear to coexist [89
, 99
, 100
]. The way AA participates in phagocytosis remains unknown, but its production seems to be important for the localized membrane exocytosis [95
] that is required for completion of phagocytosis [101
].
|
is a phosphoinositide (PI)-specific phospholipase that uses PI-4,5 bisphosphate [PI(4,5)P2] for generation of IP3 and diacylglycerol (DAG). IP3 mediates Ca2+ release from intracellular stores, and DAG is necessary for activation of several PKC isoforms (Fig. 1)
. Activation of PLC after FcR cross-linking has been reported [84
, 102
103
104
], and inhibition of this enzyme results in impaired phagocytosis in macrophages [105
] and in Dictyostelium [106
]. Also, PLC accumulates at the phagocytic cup during FcR-mediated phagocytosis [105
]. The exact role of PLC in the regulation of phagocytosis is not clear. One likely role is the activation of PKC through DAG production (Fig. 1)
, although accumulation of PLC at the phagocytic cup suggests other functions for this enzyme in phagocytosis.
PLD is an enzyme that uses phosphatidylcholine as substrate to generate choline and phosphatidic acid. PLD has been observed to become activated during phagocytosis in several systems [69
, 72
, 107
]. In neutrophils, PLD inhibition induced a decrease in the rate of phagocytosis [107
]. Also, PLD inhibition resulted in impaired PKC and Raf-1 translocation to the plasma membrane, with consequent inhibition of phagocytosis [75
]. The precise mechanism whereby PLD regulates phagocytosis has not been clearly defined. However, the phosphatidic acid (PA) generated by PLD can be converted to DAG through the action of phosphatidic acid-phosphatase-1 (PAP-1), thus making PLD activation an additional pathway leading to PKC activation [94
] (Fig. 1)
. Additionally, phosphatidic acid by itself is capable of activating various enzymes, such as PLC and PLA2 [94
] (Fig. 1)
. This notion is supported by the fact that in neutrophils, FcR-mediated degranulation is accompanied by phosphatidic acid formation, which may in turn activate PLA2 [108
].
PI-3K
PI-3K is a lipid kinase that phosphorylates the inositol ring at the 3' position [109
]. PI-3K and its lipid products, PI(3,4)P2 and PI-3,4,5-trisphosphate [PI(3,4,5)P3], are involved in a variety of signaling pathways. PI-3K products can activate some isoforms of PKC [110
, 111
] and may also be needed for local recruitment of pleckstrin homology (PH)-bearing signaling molecules such as Vav, PLC, and protein kinase B/Akt [109
, 112
]. PI-3K was initially shown to be involved in phagocytosis, when cells treated with wortmannin, a specific PI-3K inhibitor, showed reduced phagocytosis [113
]. Recently, a transient and restricted accumulation of PI-3K products has been observed at sites of phagosome formation [114
]. The main role of PI-3K during phagocytosis appears to be the regulation of pseudopod extension necessary for particle internalization. In macrophages, inhibition of PI-3K by wortmannin arrested phagocytosis at an early stage after initiation of the phagocytic signaling [101
]. Phagocytosis arrest could not be abrogated by reducing the number of particles bound to macrophages, thus indicating that reduction of plasma membrane availability was not the cause of phagocytosis arrest. However, decreasing particle size, and therefore the magnitude of pseudopod extension necessary for particle internalization, relieved the wortmannin-induced phagocytosis arrest [101
]. These results and the observation that PI-3K inhibition results in defective exocytic-membrane insertion, which leads to impaired macrophage-spreading over IgG-coated surfaces [101
], further support the notion of PI-3K as an important regulator of membrane events required for pseudopod extension (Fig. 1)
.
Additionally, it was recently reported that myosin X, an unconventional myosin with PH domains, accumulates to phagocytic cups in a wortmannin-sensitive manner and is needed for membrane-spreading on IgG-opsonized particles [115 ]. Thus, myosin X may be a molecular link among PI-3K, pseudopod extension, and particle internalization during phagocytosis.
In addition to its role in pseudopod extension, PI-3K may also regulate phagocytosis through activation of ERK (Fig. 1)
. A PI-3K-dependent pathway leading to FcR-mediated activation of ERK has been described in monocytes [61
, 79
, 105
], neutrophils [79
, 116
], and macrophages [79
].
The use of PI-3K for regulation of phagocytosis appears to be a distinctive feature of neutrophils and macrophages (Fig. 2)
. Phagocytosis of IgG-coated particles by neutrophils and macrophages is inhibited by wortmannin [46
, 78
, 101
], whereas it has no effect on phagocytosis by monocytes [61
, 79
]. PI-3K, however, is indeed activated on FcR stimulation in these cells, and its activity is necessary for cytokine production [117
] (Fig. 2)
. Moreover, we have recently found that differentiation of monocytes into macrophages involves the recruitment of PI-3K for regulation of phagocytosis [79
] (Fig. 2)
. This observation suggests that efficient phagocytosis requires regulation by PI-3K and its products (see below).
ERK
ERK is a serine/threonine kinase involved in signal transduction by a wide variety of receptors [118
]. ERK mediates activation of nuclear factors, such as Elk and nuclear factor-
B, which are important for cytokine expression [117
, 119
]. However, the role of ERK in phagocytosis is not as clear. At least two pathways leading to ERK activation on FcR stimulation have been described. In phagocytes, ERK activation may result from PKC [46
, 77
, 79
] or PI-3K activation [61
, 79
, 116
] (Fig. 1)
. The PKC-dependent pathway for ERK activation appears to involve translocation of PKC and Raf-1 to the plasma membrane [46
, 75
]. Raf-1 in turn activates MAPK kinase (MEK), and MEK activation directly leads to ERK activation [120
]. The molecules linking PI-3K to ERK activation are still to be identified, but a role for Akt in ERK activation is possible [121
]. Independently of the activation pathway, inhibition of ERK by the MEK/ERK inhibitor PD98059 abolishes phagocytosis in neutrophils and macrophages [46
, 75
, 79
, 122
, 123
]. It is interesting that ERK inhibition has no effect on phagocytosis by monocytes [61
, 79
, 87
] (Fig. 2A)
. Thus, it appears that the use of ERK for regulation of phagocytosis is a distinctive feature of neutrophils and macrophages (Fig. 2B)
.
The role of ERK in phagocytosis may be the activation of PLA2 and the production of AA (Fig. 1)
. In neutrophils and macrophages, it has been demonstrated that FcR-induced, cPLA2-mediated AA release requires ERK [99
, 100
] (Fig. 2B)
. In contrast, ERK inhibition has no effect in FcR-induced AA release in monocytes [87
]. In these cells, PKC activates iPLA2 and leads to AA production independently of ERK [89
] (Fig. 2A)
. Consistent with the restricted use of ERK for phagocytosis by neutrophils and macrophages, monocyte differentiation into macrophages involves recruitment of ERK for regulation of phagocytosis [79
] (Fig. 2)
. Moreover, during monocyte-to-macrophage differentiation, ERK and PI-3K are recruited for phagocytosis in an ordered manner. Although ERK is recruited first, fully differentiated macrophages use PI-3K and ERK for regulation of phagocytosis [79
]. The notion that these enzymes are required for efficient phagocytosis is supported by the observation that only fully differentiated macrophages achieve maximal phagocytosis on phorbol 12-myristate 13-acetate stimulation. Monocytes and partially differentiated macrophages (whose phagocytic activity is only ERK-dependent) show only a modest increase in phagocytosis on stimulation [79
]. In addition to its role in PLA2 activation, ERK may also regulate phagocytosis by modulating actin dynamics. Myosins are a large family of ATPases whose interaction with the actin cytoskeleton is thought to provide the mechanical force necessary for pulling the forming phagosome into the cytoplasm. Phosphorylation of some myosins by myosin light chain kinase (MLCK) results in myosin activation [124
] (Fig. 1)
. Inhibition of MLCK in neutrophils results in suppression of phagocytosis [123
]. As in these cells MLCK activation is ERK-dependent, ERK may also regulate particle internalization through activation of actin-binding proteins.
GTPases
The actin cytoskeleton is fundamental for phagocytosis. Members of the Rho family of small GTPases, including Rho, Rac, and Cdc42, are important in the reorganization of the actin cytoskeleton leading to formation of stress fibers, fillopodia, and lamellipodia [125
]. Participation of these enzymes in phagocytosis has been demonstrated through pharmacological inhibition and also by expression of dominant-negative forms of the enzymes [126
]. Inhibition of Rho by C3 transferase, an exoenzyme from Clostridium botulinum, which adenosine 5'-diphosphate (ADP) rybosylates and inactivates Rho [125
], resulted in impaired F-actin formation and internalization of IgG-coated particles after FcR engagement [127
]. However, C3 transferase treatment of mouse J774 macrophages did not have an effect on phagocytosis [128
, 129
]. Thus, the participation of Rho in phagocytosis remains controversial.
In contrast, participation of Rac and Cdc42 in phagocytosis is firmly established (Fig. 1) . Inhibition of either enzyme in macrophages results in complete inhibition of actin assembly at nascent phagosomes and internalization of IgG-coated particles [128 , 130 ]. Accumulation of active Cdc42 at the cytoplasmic side of the plasma membrane, beneath bound particles, triggers actin assembly and formation of finger-like extensions of membrane around the particle [131 ]. Similarly, localized Rac activation results in particle internalization, although in this case, there are not membrane extensions [132 ]. These observations suggest that Rac and Cdc42 have different roles in regulation of phagocytosis [12 ], although both enzymes appear to regulate phagocytosis through modulation of actin dynamics leading to pseudopod extension.
The mechanism of Rac and Cdc42 activation involves transition from an inactive guanosine 5'-diphosphate (GDP)-bound to an active guanosine 5'-triphosphate (GTP)-bound state. This transition is catalyzed by guanine nucleotide-exchange factors (GEFs). More than 50 GEFs for the Rho family of GTPases have been identified in the human genome [133
]. However, the way they participate in receptor signaling for activation of specific GTPases is unknown. One of these GEFs, Vav, has been implicated in phagocytosis [128
, 130
, 134
, 135
]. Vav is a multidomain protein comprising an amino terminal calponin homology domain, an acidic region, a Dbl homology domain, PH domains (common to almost all Rho GEFs), a zinc-finger motif, a proline-rich region, and a carboxy terminal SH-3–SH-2–SH-3 module [136
]. The GEF activity of Vav can be modulated by tyrosine phosphorylation [137
] and by PI lipids [138
]. A very important role of Vav in phagocytosis is indicated by recent evidence showing that Vav is recruited to sites of phagosome formation during FcR-mediated phagocytosis [134
, 135
] but not complement receptor-mediated phagocytosis [135
]. It is interesting that Rac recruitment to nascent phagosomes takes place in the absence of Vav exchange activity [135
], suggesting that Rac is recruited in its inactive, GDP-bound state to nascent phagosomes, where Vav induces its activation. Although all these reports indicate that Vav has an important role in phagocytosis, preliminary data presented at a recent meeting show that macrophages from Vav-/- knockout mice had normal, IgG-mediated phagocytosis [139
]. These data suggest that other GEFs, besides Vav, may also activate Rac during phagocytosis. The GEFs that are relevant for Cdc42 activation during phagocytosis remain to be determined.
Once localized to sites of phagocytosis, Rac and Cdc42 may exert their action through the Wiskott-Aldrich syndrome protein (WASP) [140
141
142
143
] (Fig. 3
). The Wiskott-Aldrich syndrome (WAS) is a rare, inherited X-chromosome-linked, recessive disease characterized by immune dysfunction and microthrombocytophenia [144
]. WASP, expressed exclusively on hematopoietic cells, binds directly to Cdc42 and Rac in a GTP-dependent manner [145
]. WASP is also actively recruited to the phagocytic cup during IgG-mediated phagocytosis, and phagocytes isolated from WAS patients (whose cells express little or no WASP) showed attenuated actin-cup formation and reduced phagocytosis [141
]. Also, macrophages derived from WASP-deficient mice show that WASP is necessary for efficient phagocytosis of apoptotic cells [143
]. WASP and related family proteins (N-WASP and SCAR) bind in turn via their carboxy terminus to the seven-subunit Arp2/3 complex [140
, 146
] (Fig. 3)
. This complex has been shown to be the major actin nucleator in cells [147
, 148
]. The Arp2/3 complex also accumulates at FcR- and complement receptor 3-mediated phagosomes and is needed for particle ingestion by these receptors [129
]. Other domains of WASP are required for the formation of molecular complexes that also participate in actin polymerization. The C-terminal verprolin homology-cofilin homology-acidic domain mobilizes WASP to the plasma membrane, and the proline-rich domain binds other proteins, such as VASP [149
] (Fig. 3)
. VASP is also recruited to nascent phagosomes during FcR-mediated phagocytosis [134
]. Thus, FcR trigger signaling events that lead to the recruitment of various molecules that converge to regulate actin polymerization during phagocytosis.
|
GTPases of the ARF family that modulate membrane-recycling events [157
] have also been identified as important regulators of phagocytosis. ARF6, a member of this family, is required for actin assembly and particle ingestion during FcR-mediated phagocytosis [158
]. The ARF6 protein interacts with the Rac effector POR-1 [159
] and functions downstream of Rac in actin polymerization [160
] (Fig. 3)
. In addition, the role of ARF6 in membrane recycling is important for phagocytosis. Defective membrane insertion into forming phagosomes results in partial particle enclosure [101
]. This is indeed the phagocytic phenotype observed in cells expressing a dominant-negative form of Rac [161
]. As Rac induces particle internalization only if its POR-1-interacting region is intact [132
], it is clear that ARF6 also functions downstream of Rac in membrane recycling. Thus, FcR-mediated signaling, leading to actin remodeling and membrane extension during phagocytosis, involves Rac, POR-1, and ARF6 (Fig. 3)
.
ARF6 appears to cycle between an intracellular compartment and the plasma membrane, depending on its activation state. GTP-bound ARF6 accumulates at the plasma membrane, whereas GDP-bound ARF6 localizes to endosomal vesicles [159
]. As expression of ARF6 mutants defective in GTP hydrolysis or GTP-binding results in impaired phagocytosis, it has been suggested that cycling between GTP- and GDP-bound states is important for ARF6 function in phagocytosis [158
]. This notion is stressed by the fact that PAG3, a GTPase-activating protein for ARF6, accumulates with ARF6 and F-actin at phagocytic cups during FcR-mediated phagocytosis [162
]. Also, alteration of the ARF6-GTP/ARF6-GDP balance through PAG3 overexpression reduced F-actin levels at phagosomes and phagocytosis efficiency [162
]. Thus, FcR trigger signaling events that lead to the recruitment of various molecules that converge to regulate actin polymerization and membrane recycling during phagocytosis.
Other Actin-Binding Molecules
In addition to GTPases, several other actin-binding molecules have been localized around phagosomes during FcR-mediated phagocytosis. These molecules include talin, -actinin, vinculin, gelsolin, coronin, cofilin, paxillin, and L-plastin [3
, 10
, 12
]. These molecules regulate actin dynamics through various mechanisms, including nucleation of actin, cross-linking, and stabilization of actin filaments, and anchorage of actin fibers to the membrane [3
, 12
]. However, the exact role of these proteins during phagocytosis remains unknown.
Phosphatases
In addition to activating signals, inhibitory signals help to control the level of the phagocytic response. Regulation of immune responses initiated by ITAM-containing receptors in many cell types has been shown to be dependent on the participation of inhibitory receptor systems that maintain an equilibrium between activation and inhibition signals [163
]. Inhibitory receptors signal through immunoreceptor tyrosine-based inhibition motifs (ITIMs) [164
]. The first ITIM motif was identified in a 13 amino acid sequence within the intracytoplasmic tail of the FcRIIB in B lymphocytes [165
]. This sequence is necessary and sufficient to inhibit B cell activation. When an antigen is recognized by the B cell antigen receptor (BCR), which contains ITAMs, B lymphocytes proliferate. However, in later stages of the immune response, the BCR is coaggregated with the FcRIIB as a result of the presence of antibodies against the same antigen. Coaggregation of these receptors leads to arrest of B cell activation [165
]. FcRIIB-mediated, negative regulation was later shown to function on other ITAM-containing receptors, including the TCR and the high-affinity FcRI [166
, 167
]. These findings were then confirmed in vivo with the help of FcRIIB knockout mice. These animals exhibit enhanced antibody responses [168
], exaggerated IgE- [169
] and IgG-dependent anaphylactic reactions [168
], enhanced susceptibility to IgG-dependent autoimmune diseases [170
, 171
], and enhanced ADCC to tumor antigens after injection of therapeutic antibodies [172
].
ITIM sequences have also been identified in other inhibitory receptors including killer cell inhibitory receptors, paired Ig-like receptors-B, platelet endothelial cell adhesion molecule-1/CD31, and others [163
]. All of these receptors present a general mechanism of action. When inhibitory receptors are coaggregated with activating receptors, their ITIMs are phosphorylated on tyrosines by Src family kinases. Phosphorylated ITIMs then become docking sites for protein tyrosine phosphatases (SHP)-1 and -2. SHP-1 and SHP-2 dephosphorylate tyrosines on receptors and also on effector molecules whose tyrosil phosphorylation is critical for activation. As a result of the inhibition of early signaling events, downstream signals are not generated, and cell activation is arrested [173
]. It is interesting that FcRIIB presents an alternative mechanism of inhibition. Instead of SHPs, the ITIM in this receptor recruits the Src homology 2 domain-containing inositol 5'-phosphatases (SHIP)-1 [174
, 175
] and SHIP-2 [176
]. These phosphatases act on PI(3,4,5)P3, which can allow for membrane recruitment of molecules bearing a PH domain. One of these molecules is Bruton’s tyrosine kinase, which is necessary for PLC activation and induction of a Ca2+ rise [177
]. Contrary to other ITIM-containing receptors, the selective use of SHIP by FcRIIB allows this receptor to stop the intracellular propagation of PI-3K-dependent, downstream signals without preventing early activation signals (Fig. 4
). The selective use of SHIP by FcRIIB seems to be determined by particular hydrophobic amino acid residues at the Y + 2 position within the ITIM [178
] and also by the level of tyrosine phosphorylation within the ITIM [179
].
|
R-mediated phagocytosis seems to be down-regulated by ITIM-recruited phosphatases. Overexpression of SHIP in macrophages resulted in inhibition of FcR-mediated phagocytosis [180
]. SHIP was also found accumulated at FcR and complement receptor phagocytic cups [180
]. In addition, cross-linking of FcRIIA and FcRIIB at the same time resulted in reduced phagocytosis efficiency in monocytes [18
]. The FcRIIB-induced inhibition was clearly mediated by SHIP [18
] (Fig. 4)
. More recently, the protein-tyrosine phosphatase MEG2 was found to be localized to newly formed phagosomes in neutrophils [181
]. Our knowledge of the role of these and other phosphatases in the regulation of phagocytosis is still very limited, but it is very likely that phosphatases act at late stages of this process to "turn-off" phagocytic signaling. |
ENCLOSURE OF THE PHAGOCYTIC VACUOLE |
|---|
|
TOPABSTRACTINTRODUCTIONFcR for IgG (Fc{gamma}R)SIGNAL TRANSDUCTION DURING...ENCLOSURE OF THE PHAGOCYTIC...PHAGOSOME MATURATIONCONCLUSIONREFERENCES |
|---|
R on phagocytes mediate internalization of antibody-antigen complexes and large antibody-opsonized particles. Soluble immune complexes are taken up by clathrin- and ubiquitin-dependent endocytosis, and large particles are internalized by F-actin-dependent phagocytosis [182
]. Although these two processes have fundamental differences, recent reports indicate that they share several protein components, including amphiphysin II and dynamin II [183
, 184
]. Also, phagosomes and endocytic vesicles follow similar maturation processes after internalization, as indicated by the appearance of markers for early endosomes, late endosomes, and then lysosomes [185
], as we describe next. Particle internalization by phagocytes involves not only reorganization of the actin cytoskeleton but also membrane fusion events to complete the phagocytic vacuole. Originally, it was thought that reorganization of the actin cytoskeleton pushed the cell membrane around the particle to be internalized. However, the ability of macrophages to ingest multiple particles with a total surface area actually larger than the cell itself indicated that phagocytes must have an internal source of membrane that accounts for the membrane necessary for phagosome formation [10 , 186 ]. Ultrastructural studies confirmed that membrane from an endosomal compartment is recruited to sites of phagocytosis in a PLA2-dependent manner [88 ]. PI-3K is also important for this process, as inhibition of this enzyme does not seem to affect the rate of actin polymerization [101 , 187 , 188 ], but it impedes completion of phagocytosis [161 , 189 ] and blocks exocytic insertion of membrane during phagocytosis [101 ]. Recently, this idea has been confirmed by the demonstration of the appearance of endosomal markers at the plasma membrane during phagocytosis [190 , 191 ].
Soluble NSF attachment protein receptor (SNARE) proteins play a fundamental role in membrane fusion events. Fusion of two lipid bilayers depends on the formation of specific complexes between v-SNAREs on vesicles and cognate t-SNAREs on target membranes [192 ]. Inactivation of the SNAREs VAMP2 and VAMP3 by tetanus or botulinum toxins [193 ] and of N-ethylmaleimide-sensitive factor (NSF), an ATPase important for SNARE function, by dominant-negative NSF [194 ] reduced phagocytosis efficiency. VAMP3 is predominantly localized at the recycling compartment of early endosomes [190 ]. These observations further support the notion that membrane fusion events between the plasma membrane and an endosomal compartment are required for phagocytosis (Fig. 5 ). Although membrane exocytosis could occur randomly anywhere along the plasma membrane, a chimeric protein formed by the SNARE VAMP3 and green fluorescent protein (VAMP3-GFP) translocates from recycling endosomes specifically to sites of phagosome formation [190 , 195 ]. This suggests that vesicle exocytosis is a focal event occurring at sites of phagocytosis.
|
Dynamin 2 and amphiphysin IIm are proteins associated with endocytic vesicles, which have also been reported to participate in phagocytosis [183 , 184 , 190 , 195 ] (Fig. 5) . Dynamin 2 is a GTPase required for scission of clathrin-coated endocytic vesicles from the plasma membrane and is recruited to clathrin-coated pits through its interaction with an SH-3 domain on amphiphysin IIm [184 , 199 , 200 ]. Expression of dominant-negative forms of dynamin 2 blocked pseudopod extension and arrested phagocytosis by macrophages [183 ]. Also, expression of a mutant form of amphiphysin IIm, lacking its dynamin-binding domain, arrested macrophage phagocytosis [184 ]. The phagocytic phenotype of these cells was similar to that observed in PI-3K-inhibited macrophages [101 ], and PI-3K was needed for recruitment of amphiphysin IIm and dynamin 2 to forming phagosomes [184 ]. Thus, dynamin 2 recruitment to forming phagosomes, via amphiphysin IIm, is activated by PI-3K. However, it is not clear whether dynamin 2 acts as a mechanical force-generator for vesicle scission or whether it acts as a classic GTPase, activating a downstream effector to accomplish its function [184 ].
In addition to the membrane remodeling for pseudopod extension, particle internalization requires a driving force pulling the forming phagosome into the cytoplasm. It is thought that actin polymerization is the main force driving particle internalization. However, myosins, motor-proteins that couple their ATPase activity to movement along actin fibers [201
, 202
], are also likely to provide the contractile force necessary for particle internalization (Fig. 5)
. Myosin localization around phagosomes in macrophages [203
] and neutrophils [204
] has been known for many years. Myosin I was found in phagosomes of zymosan-coated particles [205
], whereas the filamentous myosin II (or muscle myosin) is important for FcR-mediated phagocytosis by macrophages [206
] and by neutrophils [123
]. Recently, it was reported that myosin X is also recruited to phagocytic cups in a PI-3K-dependent manner [115
]. Other myosins, including myosin IC, myosin V, and myosin IXb, are also found at phagosomes [206
]. So, it is likely that at least one of these myosin isoforms is involved in generating the contractile force necessary for particle internalization. Myosin II is the best candidate for this function, as it is activated during FcR-mediated phagocytosis by the enzyme MLCK [123
]. MLCK is in turn activated by ERK [123
] (Fig. 1)
, which is known to be required for phagocytosis by neutrophils and macrophages [79
]. In Dictyostelium, myosins IB, IC, IK, II, and VII seem to be important for phagocytosis [207
208
209
210
]. This suggests that various aspects of phagocytosis are probably mediated by particular myosin isoforms. This idea will certainly be investigated further in the future.
|
PHAGOSOME MATURATION |
|---|
|
TOPABSTRACTINTRODUCTIONFcR for IgG (Fc{gamma}R)SIGNAL TRANSDUCTION DURING...ENCLOSURE OF THE PHAGOCYTIC...PHAGOSOME MATURATIONCONCLUSIONREFERENCES |
|---|
Phosphoinositide distribution during phagocytosis has been visualized by the use of protein probes consisting of the GFP fused to the plekstrin homology domains that bind to different PI products of PI-3K [211
212
213
]. These probes have shown that various PI-3K products rapidly and transiently accumulate at sites of phagocytosis upon FcR engagement [105
, 114
, 214
, 215
]. PI(4,5)P2 [105
] and PI(3,4,5)P3 [114
] are formed exclusively at the phagocytic cup and rapidly disappear as the phagosome seals and scisses form the plasma membrane (Fig. 6
). A high level of PI 3-phosphate [PI(3)P] is then formed in the phagosome membrane immediately after sealing from the plasma membrane and remains for several minutes [214
, 215
]. PI(4,5)P2 is produced by phosphorylation of PI(4)P by the enzyme PI(4)P 5-K, which is also found at the phagocytic cup [105
]. It is interesting that when levels of PI(4,5)P2 go down, mobilization of PLC and accumulation of DAG are also observed [105
] (Fig. 6)
. PI(3,4,5)P3 is produced by phosphorylation of PI(4,5)P2 by PI-3K, which is also activated during phagocytosis [46
, 78
, 79
, 101
] and accumulates at the phagocytic cup [114
] (Fig. 6)
. As PI(3,4,5)P3 and PI(3)P are the products of different classes of PI-3K [211
], the involvement of these classes of PI-3K in phagocytosis was investigated directly. Using cells with the regulatory subunits of class I PI-3K deleted, it was found that phagocytosis of small particles (<3 µm) proceeds to complete phagosomes. In contrast, phagocytosis of larger particles is severely impaired [214
]. These results suggest that class I PI-3K is needed during the internalization step but not for phagosome maturation. Using cells microinjected with an inhibitory antibody against class III PI-3K, it was found that particle internalization is not affected, whereas phagosome maturation is inhibited [214
, 216
]. Thus, there is a sequential production and disappearance of PI(3,4,5)P3 and PI(3)P during formation and maturation of phagosomes, consistent with the respective roles of class I and class III PI-3K in phagocytosis (Fig. 6)
.
|
, and Akt [112
]. The role of PI(3)P during phagosome maturation may be to recruit proteins for targeting vesicles. One such protein is the early endosome autoantigen (EEA1), which interacts with Rab5 and with PI(3)P to mediate endosome-endosome fusion [222
223
224
] (Fig. 6)
. Rab5 accumulates in newly formed phagosomes of macrophages ingesting Escherichia coli [225
], and EEA1 inhibition by microinjection of antibodies blocks phagosome maturation [216
]. EEA1 also binds to syntaxin-6, a protein involved in fusion of Golgi complex-derived vesicles with early endosomes and phagosomes [226
]. Thus, PI(3)P will recruit EEA1 to direct the acquisition of molecules (such as Rab5) from internal vesicles to the phagosome (Fig. 6)
. It is interesting to realize that some microorganisms, such as Mycobacterium tuberculosis, are able to survive inside phagosomes by interfering with this maturation process. Mycobacterium phagosomes contain Rab5 but fail to recruit EEA1 and syntaxin-6 [216 ]. In addition, phagosomes containing latex beads coated with lipoarabinomannan, a GPI produced by M. tuberculosis, also present reduced EEA1 recruitment [216 ]. The mechanism by which lipoarabinomannan inhibits EEA1 recruitment is currently unknown. One possibility is that it interferes with the production of PI(3)P [212 , 216 ].
More mature phagosomes accumulate other molecules such as Rab7 [227
] and Flotillin-1 [228
]. Rab7-containing vesicles appear to route material from Rab5-containing vesicles to lysosomes [225
]. Thus, it is possible that sequential association of Rab5 and Rab7 to mature phagosomes is a prerequisite for phagosome fusion with lysosomes to form the phagolysosome, where the ingested particle is ultimately destroyed (Fig. 6)
. A very recent report indicates that the cytoplasmic tail of FcRIIA plays an important role in phagolysosome formation [229
]. Wild-type FcRIIA supported phagolysosome formation, whereas tail-minus FcRIIA did not. Also, an FcRIIA with its ITAM mutated was still able to support phagolysosome formation [229
]. Thus, the cytoplasmic tail of FcRIIA contributes to phagolysosome fusion by a mechanism that does not require a functional ITAM. Whether the cytoplasmic tails of the other FcR have a similar role on regulation of phagolysosome formation is unknown.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTION |
|---|
